the e3 ubiquitin ligase trim31 attenuates nlrp3 inflammasome … · 2019-09-17 · ubiquitin ligase...

ARTICLE

Received 2 Jun 2016 | Accepted 28 Oct 2016 | Published 8 Dec 2016

The E3 ubiquitin ligase TRIM31 attenuates NLRP3inflammasome activation by promotingproteasomal degradation of NLRP3Hui Song1,2,*, Bingyu Liu1,2,*, Wanwan Huai1,2,*, Zhongxia Yu1,2, Wenwen Wang1,2, Jing Zhao1,2, Lihui Han1,2,

Guosheng Jiang3, Lining Zhang1,2, Chengjiang Gao1,2 & Wei Zhao1,2

The NLRP3 inflammasome has a fundamental role in host defence against microbial

pathogens and its deregulation may cause diverse inflammatory diseases. NLRP3 protein

expression is a rate-limiting step for inflammasome activation, thus its expression must be

tightly controlled to maintain immune homeostasis and avoid detrimental effects. However,

how NLRP3 expression is regulated remains largely unknown. In this study, we identify E3

ubiquitin ligase TRIM31 as a feedback suppressor of NLRP3 inflammasome. TRIM31 directly

binds to NLRP3, promotes K48-linked polyubiquitination and proteasomal degradation of

NLRP3. Consequently, TRIM31 deficiency enhances NLRP3 inflammasome activation and

aggravates alum-induced peritonitis in vivo. Furthermore, TRIM31 deficiency attenuates the

severity of dextran sodium sulfate (DSS)-induced colitis, an inflammatory bowel diseases

model in which NLRP3 possesses protective roles. Thus, our research describes a mechanism

by which TRIM31 limits NLRP3 inflammasome activity under physiological conditions and

suggests TRIM31 as a potential therapeutic target for the intervention of NLRP3 inflamma-

some related diseases.

DOI: 10.1038/ncomms13727 OPEN

1 Department of Immunology, Shandong University School of Medicine, Jinan, Shandong 250012, China. 2 Key Laboratory of Infection and Immunity ofShandong Province, Shandong University School of Medicine, Jinan, Shandong 250012, China. 3 Institute of Basic Medicine, Shandong Academy of MedicalSciences, Jingshi Road 18877, Jinan, Shandong 250062, China. * These authors contributed equally to this work. Correspondence and requests for materialsshould be addressed to C.G. (email: [email protected]) or to W.Z. (email: [email protected]).

NATURE COMMUNICATIONS | 7:13727 | DOI: 10.1038/ncomms13727 | www.nature.com/naturecommunications 1

mailto:[email protected]

mailto:[email protected]

http://www.nature.com/naturecommunications

NLRP3 inflammasome is a multi-protein platform whichcomprises NLRP3, ASC and caspase-1, and plays crucialroles in host defence against pathogens1–5. Inflammasome

complex assembly is triggered by stimuli from both microbialinfection and endogenous ‘danger signal’, such as nigericin,crystals, extracellular ATP, amyloid-b and alum and so on. Andthen the inflammasome complex serves as platforms for theactivation of the cysteine protease caspase-1, which cleavedpro-IL-1b and pro-IL-18 into mature IL-1b and IL-18 (refs 1–5).NLRP3 inflammasome has been implicated in many kinds ofdiseases, such as gout, autoimmune disorders, atherosclerosis,type 2 diabetes and obesity5–9. Thus, NLRP3 inflammasomeactivity must be tightly controlled to maintain immunehomeostasis and avoid detrimental effects.

NLRP3 expression level is considered as a limiting step ininflammasome activation10,11. In resting macrophages, theprotein level of NLRP3 is relatively low, so that NLRP3inflammasome assembly is hardly induced11,12. Induction ofNLRP3 protein expression licensed by TLR ligands (signal 1)allows respective NLRP3 activators (signal 2) to trigger caspase-1cleavage10,13. Thus, regulation of NLRP3 level offers aninteresting mechanism to alter the inflammatory potential ofimmune cells11. Up to now, several mechanisms for negativeregulation of NLRP3 expression have been established. Forexample, we reported that aryl hydrocarbon receptor could bindto the NLRP3 promoter and inhibit its expression attranscriptional level14. MiR-223 suppresses NLRP3 mRNAexpression through a conserved binding site within the 30

untranslated region of NLRP3 (refs 15,16).Autophagy–lysosomal pathway and ubiquitin–proteasome

pathway are two major protein degradation systems widely existin mammalian cells, which provide specificity and regulate theintensity of innate immune responses. It has been reportedthat autophagy-dependent degradation was involved in theregulation of NLRP3 expression17,18. For example, plasminogenactivator inhibitor type 2 enhances NLRP3 degradation in anautophagy-dependent manner17. Dopamine D1 receptor (DRD1)signalling promotes NLRP3 ubiquitination via E3 ubiquitin ligaseMARCH7, leading to the autophagy-mediated degradation ofNLRP3 (ref. 18). However, whether ubiquitin–proteasomepathway is also involved in the regulation of NLRP3 proteinexpression remains unknown.

TRIM family proteins have been implicated in the negativeregulation of innate immune responses by promoting degradationof their respective substrates through ubiquitin–proteasomepathway19. TRIM31 is a member of the TRIM protein family,which is encoded within the major histocompatibility complex(MHC) class I region20. Several TRIM family members (includingTRIM31, TRIM40, TRIM10, TRIM15, TRIM26 and TRIM39) areorganized in a tight cluster and two TRIM genes (includingTRIM38 and TRIM27) telomeric of the cluster within the MHCclass I region20. Most of the above mentioned TRIM familymembers possess vital regulatory effects in innate immuneresponses19,21–27. However, the biological function of TRIM31,especially in the innate immune response, remains unknown.

In this report, we describe a mechanism by which TRIM31negatively regulates NLRP3 inflammasome activity. TRIM31could directly bind to NLRP3, and then promote K48-linkedpolyubiquitination and proteasomal degradation of NLRP3 inboth resting and activated macrophages. The identification ofTRIM31 as a physiological suppressor of NLRP3 expressionprovides an explanation on the limitation of NLRP3 inflamma-some activation under physiological conditions. Consistent withthat, TRIM31 deficiency facilitated NLRP3 inflammasomeactivation, enhanced IL-1b secretion and thus aggravatedalum-induced peritonitis in vivo. On the other hand, TRIM31

deficiency attenuated the severity of dextran sodium sulfate(DSS)-induced colitis by promoting NLRP3 inflammasomeactivation, which was reported to be possessing protectiveroles in the model. Thus, TRIM31 could be a potentialtherapeutic target for NLRP3-associated syndromes, includingautoinflammatory and autoimmune diseases.

ResultsTRIM31 specifically inhibits NLRP3 inflammasome activation.To investigate the potential functions of TRIM31 in innateimmune response, specific and effective siRNA targeting TRIM31was used to suppress the expression of endogenous TRIM31 inmouse macrophages (Supplementary Fig. 1a). TRIM31 knock-down has no regulatory effects on mRNA expression oflipopolysaccharide (LPS)-induced proinflammatory cytokines,including TNF-a and IL-1b (Supplementary Fig. 1b). But, IL-1bsecretion was significantly increased in TRIM31 silenced mac-rophages primed by LPS and then treated by NLRP3 inflamma-some activator such as ATP, nigericin or alum (Fig. 1a). However,TNF-a and IL-6 secretion were not influenced by TRIM31knockdown (Fig. 1a). Similarly, TRIM31 overexpression greatlyinhibited IL-1b secretion in human THP-1 cells (Fig. 1b). Toconfirm the inhibitory role of TRIM31 in IL-1b secretion, theeffects of TRIM31 deficiency on IL-1b expression in mouseperitoneal macrophages were observed. ATP and Nig stimulatedIL-1b secretion by LPS-primed TRIM31-deficient macrophageswas significantly increased (Fig. 1c, Supplementary Fig. 1c–e).However, IL-1b secretion stimulated by poly(dA:dT) (an AIM2inflammasome activator) and flagellin(a NLRC4 inflammasomeactivator) was not influenced by TRIM31 deficiency (Fig. 1c,Supplementary Fig. 1d,e). In addition, TRIM31 deficiencyalso enhanced ATP stimulated IL-1b secretion in PGN-primedmacrophages (Supplementary Fig. 1f). Furthermore, TRIM31deficiency also had no inhibitory effects on TNF-a, IL-6expression and IL-1b mRNA expression (Supplementary Fig. 1g).Taken together, these data indicated that TRIM31 specificallyinhibited NLRP3 inflammasome activation.

Caspase-1 cleavage is a critical step for the NLRP3 inflamma-some activation. We then investigated the effects of TRIM31 oncaspase-1 cleavage. TRIM31 knockdown and deficiency bothsignificantly enhanced caspase-1 cleavage in LPS-primed mouseperitoneal macrophages treated by ATP (Fig. 1d,e). Accordingly,less cleaved caspase-1 was detected in TRIM31 overexpressedTHP-1 cells (Fig. 1f). However, the TRIM31 truncated mutantDRING, in which the N-terminal RING domain was deleted andthe potential E3 ubiquitin ligase activity was deprived, lost theability to inhibit caspase-1 cleavage, compared with TRIM31 WT(Fig. 1f).

To further rule out the possible regulatory effects of TRIM31 inthe priming process of NLRP3 inflammasome activation, weexplored the function of TRIM31 on NF-kB and MAPKactivation. TRIM31 knockdown had no influence on LPS-inducedphosphorylation of IkB-a, JNK, ERK and p38 (SupplementaryFig. 1h). Additionally, no differences of MyD88- and TNF-a-induced NF-kB reporter gene activation were observed inTRIM31 overexpressed HEK293T cells (Supplementary Fig. 1i).Collectively, these data indicated that TRIM31 specificallyinhibited NLRP3 inflammasome activation and subsequentIL-1b secretion.

TRIM31 promotes proteasomal degradation of NLRP3. TRIMfamily proteins have been implicated in the negative regulation ofinnate immune responses by promoting degradation of theirrespective substrates through ubiquitin–proteasome pathway19.We next sought to determine whether TRIM31 could inhibit

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms13727

2 NATURE COMMUNICATIONS | 7:13727 | DOI: 10.1038/ncomms13727 | www.nature.com/naturecommunications


NLRP3 inflammasome activation by promoting proteindegradation of inflammasome components. TRIM31knockdown greatly increased NLRP3 protein expression inresting mouse peritoneal macrophages, with no influence onASC and caspase-1 protein expression (Fig. 2a). Similarly,TRIM31 knockdown also significantly increased NLRP3 proteinlevel in LPS-priming macrophages (Fig. 2b). But, TRIM31knockdown had no effects on NLRP3 mRNA expression(Fig. 2a,b). Consistently, TRIM31 overexpression greatlydecreased NLRP3 protein expression in a dose-dependentmanner in HEK293T cells (Fig. 2c). Importantly, similar resultswere also observed in TRIM31 deficient mouse peritonealmacrophages. TRIM31 deficiency greatly enhanced NLRP3protein expression in both resting and activated macrophages(Fig. 2d, Supplementary Fig. 2). But, TRIM31 deficiency could notincrease AIM2 and NLRC4 expression (Fig. 2d). We nextexamined the effects of TRIM31 deficiency on other agonists-induced NLRP3 expression. TRIM31 deficiency greatly enhancedIL-1b-, poly(I:C)- and PGN-induced NLRP3 expression inmacrophages (Fig. 2e). However, no marked difference of TNF-a-induced NLRP3 expression was observed in TRIM31 deficientmacrophages (Fig. 2e).

Given the functions of TRIM family members in promotingthe degradation of their target proteins, we inferred thatTRIM31 might inhibit NLRP3 protein expression via promotingits degradation. TRIM31 deficiency inhibited NLRP3 proteindegradation (Fig. 2f,g). However, no differences of ASC,

Caspase-1 and NLRC4 protein degradation were observed(Fig. 2f,g). These data indicated that TRIM31 inhibited NLRP3expression via promoting its protein degradation. We thenexplored the degradation pathway of NLRP3 mediated byTRIM31. TRIM31-induced NLRP3 degradation could be reversedby proteasome inhibitor MG-132, but not by lysosome inhibitorchloroquine (Fig. 2h) or autophagy inhibitor 3-MA (Fig. 2i),indicating that TRIM31 mediated NLRP3 degradation inproteasome. Next, we examined the NLRP3 expression in bothwild type (WT) and TRIM31-deficient macrophages followingMG132 treatment. MG132 treatment slightly enhanced NLRP3expression in unactivated WT macrophages, with no effects onLPS-primed WT macrophages (Fig. 2j). In MG132-treatedmacrophages, TRIM31 deficiency could not enhance NLRP3expression (Fig. 2j). Interestingly, TRIM31 deficiency greatlyinhibited NLRP3 expression in LPS-primed and MG132-treatedmacrophages (Fig. 2j). These data indicated that MG132, as ageneral proteosome inhibitor, could not markedly affect NLRP3expression in WT macrophages. However, MG132 treatmentcould block or reverse the effects of TRIM31 deficiency onNLRP3 expression. Taken together, these results indicated thatTRIM31 could promote proteasomal degradation of NLRP3.

TRIM31 interacts with NLRP3. To further investigate howTRIM31 promotes NLRP3 degradation, we first examined theinteraction between TRIM31 and NLRP3. In vitro binding assays

1,600a b

c

d e f

**

** ** **

**

1,200

800

IL-1

β (p

g m

l–1)

IL-1

β (p

g m

l–1)

IL-1

β (p

g m

l–1)

IL-6

(ng

ml–1

)

IL-6

(ng

ml–1

)

TN

F-α

(ng

ml–1

)

TN

F-α

(ng

ml–1

)

400

0

0

LPS LPS (h)––

––+

+ ++

Ctrl siRNA TRIM31 siRNA

ATP ATP0– –+ + + +

04 8 84

Caspase-1 p20 Caspase-1 p20

Caspase-1 Caspase-1

25 kDa 25 kDaSN SN

CL

CL

SN

CL

15 kDa 15 kDa55 kDa40 kDa 40 kDa

25 kDa

100 kDa40 kDa

40 kDa

25 kDa

15 kDa

25 kDa

40 kDa

40 kDa

Actin

Actin

IL-1β p17

LPS

Ctrl

– ––

–––

+ ++

+++

WT ΔRING

ATPCaspase-1 p20

Caspase-1

Actin

IL-1β p17IL-1β p17

IL-1β p31 IL-1β p31

NLRP3

100200300400500600 **

**

0

0

0 0

300

600

900Ctrl siRNA Ctrl

TRIM31 siRNA TRIM31

TRIM31+/+

TRIM31+/+

TRIM31–/–

TRIM31–/–

0.51

1.52

0

0.5

1

1.5

2

2.53

3

3

6

6

9

9

12

15

PBS

PBSPBS

PBS

PBSLP

S

LPS

LPS

LPS

LPS

LPS+AT

P

LPS+AT

P

LPS+AT

P

LPS+AT

P

LPS+AT

P

LPS+Nig

LPS+Nig

LPS+Nig

LPS+Nig

LPS+po

ly(dA

:dT)

LPS+po

ly(dA

:dT)

LPS+po

ly(dA

:dT)

LPS+fla

gellin

LPS+fla

gellin

LPS+fla

gellin

LPS+Nig

LPS+alu

m

LPS+alu

mPBS

PBSLP

S

LPS+AT

P

LPS 4h

+ATP

LPS 8h

+ATP

LPS+Nig

LPS+alu

m

Figure 1 | TRIM31 specifically inhibits NLRP3 inflammasome activation. (a) ELISA of IL-1b, TNF-a and IL-6 in supernatants from mouse peritoneal

macrophages silenced of TRIM31, primed with LPS for 8 h, and followed by stimulation with ATP, Nig. or alum for 30 min. (b) ELISA of IL-1b in supernatants

from THP-1 cells transfected with TRIM31 plasmid, primed with LPS for various times, and followed by stimulation with ATP for 30 min. (c) ELISA of IL-1b,

TNF-a and IL-6 in supernatants of mouse peritoneal macrophages from TRIM31þ /þ or TRIM31� /� mice, primed with LPS for 8 h, and followed by

stimulation with ATP, Nig., poly(dA:dT) or flagellin for 30 min. (d) Immunoblot analysis of supernatants (SN) or cell lysates (CL) from mouse peritoneal

macrophages silenced of TRIM31, primed with LPS, and followed by stimulation with ATP for 30 min. (e) Immunoblot analysis of supernatants (SN) or cell

lysates (CL) of mouse peritoneal macrophages from TRIM31þ /þ or TRIM31� /� mice, primed with LPS, and followed by stimulation with ATP for 30 min.

(f) Immunoblot analysis of supernatants (SN) or cell lysates (CL) from THP-1 transfected with TRIM31 WT or DRING mutant, primed with LPS, and

followed by stimulation with ATP for 30 min. Similar results were obtained in three independent experiments. **Po0.01. (Student’s t-test). Data are

representative of three experiments (mean and s.d. of six samples in a–c).

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms13727 ARTICLE



demonstrated that TRIM31 could directly interact with NLRP3(Fig. 3a). And then, to confirm the interaction between TRIM31and NLRP3 in vivo, we assessed resting and LPS-stimulatedmacrophages by immunoprecipitation (IP). An associationbetween TRIM31 and NLRP3 could be detected in both mouseperitoneal macrophages and THP-1 cells (Fig. 3b, SupplementaryFig. 3a). To investigate whether TRIM31 interacted with otherNLRP3 inflammasome molecules, we expressed TRIM31 withNLRP3, caspase-1 and ASC respectively, in HEK293T cells.TRIM31 could immunoprecipitate with NLRP3, and could notimmunoprecipitate with caspase-1 and ASC (Fig. 3c,Supplementary Fig. 3b). Consistently, confocal analysis alsodemonstrated the colocalization between TRIM31 and NLRP3(Fig. 3d, Supplementary Fig. 3c). Taken together, these datasuggested that TRIM31 could directly interact with NLRP3.

TRIM31, as a member of the TRIM family proteins, containsan N-terminal RING-finger domain, B-box domain and twoC-terminal coiled-coil (CC) domains (Fig. 3e). To search for thedomains of TRIM31 that is responsible for the interaction withNLRP3, a series of Flag-tagged TRIM31 truncated mutants wereconstructed (Fig. 3e). NLRP3 was coprecipitated with TRIM31WT, RING domain deletion mutant (DRING), C126 and C270,but not with N131 and N162 (Fig. 3f). These results indicatedthat TRIM31 interacted with NLRP3 via its second C-terminalcoiled-coil domain. Three truncated mutants of NLRP3 were alsoconstructed (Fig. 3g). Co-IP experiments with these truncatedmutants showed that NLRP3 DPYD lost the ability to bind

TRIM31, while NLRP3 DLRR and DNACHT did not (Fig. 3h).So the PYD domain of NLRP3 is required for theNLRP3–TRIM31 interaction.

TRIM31 promotes K48-linked polyubiquitination of NLRP3.Protein ubiquitination is a key step in the ubquitin–proteasomedegradation pathway28. NLRP3 could be ubiquitinated with bothK48 and K63 linkage in macrophages (Supplementary Fig. 4), andthat is consistent with a previous report18. We have identified E3ubiquitin ligase TRIM31 as a NLRP3-associated protein (Fig. 3)and that promoted us to investigate whether TRIM31 couldmediate the ubiquitination of NLRP3. NLRP3 was cotransfectedwith HA-ubiquitin and WT TRIM31 into HEK293T cells. NLRP3ubiquitination was markedly increased in the presence ofTRIM31 expression plasmid (Fig. 4a). Importantly, the TRIM31point mutation (C16AC36A) with substitution of the cysteineresidue with alanine at position 16 and 36 within the RINGdomain, lost the ability to promote polyubiquitination of NLRP3(Fig. 4a), indicating TRIM31 could promote the ubiquitination ofNLRP3 though the N-terminal RING-finger domain. To studythe forms of TRIM31-mediated NLRP3 polyubiquitination,ubiquitin mutant vectors K48 and K63, which containarginine substitutions of all of its lysine residues except the oneat position 48 and 63 respectively, were used in the transfectionassays. TRIM31 promoted NLRP3 polyubiquitination could bedetected in the presence of K48 plasmid, but not with K63

TRIM31siRNA

a b d e

c

f

h i j

g

– +

100 kDa100 kDa

100 kDa

100 kDa

100 kDa100 kDa

100 kDa

100 kDa

100 kDa

100 kDaMyc Myc

MG1323-MA (mM)Chloroquine

Myc-NLRP3 + + + ++ + + + + +

+

+ + + +

++++

++

+–

––– – – –

– – – –0 05 510 10––– Myc-NLRP3

100 kDa

LPS (h)LPS (h)

Ctrl siRNA

0 000

TRIM31+/+

TRIM31+/+

TRIM31+/+

TRIM31+/+

TRIM31+/+

TRIM31–/–

TRIM31–/–

TRIM31–/–

TRIM31–/–

TRIM31–/–

4 444

8 888

0 04 48 812 12 1.2

**

**

1.2

1 1

0.8

0.6

0.4

0.2

0.8

0.6

0.4

0.2

0 0

1.2

1

0.8

0.6

0.4

0.2

0

1.2

1

0.8

0.6

0.4

0.2

0

0 4 8 12 0 4 8 12 0 4 8 12 0 4 8 12CHX (h)

TRIM31 siRNA

NLRP3

NLRP3

NLRP3

NLRP3

NLRP3

PBS

WT WT WT WT WT WTKO KO KO KO KO KO

PGNLPSPoly(I:C)IL-1β TNF-α

25 kDa25 kDa

25 kDa

40 kDa40 kDa

40 kDa40 kDa

40 kDa

40 kDa

40 kDa

40 kDa

40 kDa

40 kDa

40 kDa

40 kDa 40 kDa40 kDa 40 kDa

40 kDa

NLRP3

NLRP3

CHX (h)

ASC

ASC

ASC

ASC

Caspase-1

Caspase-1

Caspase-1

Caspase-1

Actin

Actin

LPS

DMSO MG132

NLRP3

Actin

Actin

Actin

Actin Actin

Flag

Flag Flag

Flag-TRIM31Flag-TRIM31

Flag-TRIM31

Actin

NLRC4

NLRC4

NLRC4

AIM2

Actin

NLRP3mRNA

NLRP3mRNA

ActinmRNA

ActinmRNA

–

Figure 2 | TRIM31 promotes proteasomal degradation of NLRP3. (a) Immunoblot analysis of extracts (upper panel) or RT-PCR analysis (lower panel) of

mouse peritoneal macrophages silenced of TRIM31. (b) Immunoblot analysis of extracts (upper panel) or RT-PCR analysis (lower panel) of mouse

peritoneal macrophages silenced of TRIM31, then stimulated for various times with LPS. (c) Immunoblot analysis of extracts from HEK293T cells

transfected with increasing amount of TRIM31 expression plasmid. (d) Immunoblot analysis of extracts from TRIM31þ /þ or TRIM31� /� mouse

peritoneal macrophages, then stimulated for various times with LPS. (e) Immunoblot analysis of NLRP3 expression from TRIM31þ /þ or TRIM31� /�

mouse peritoneal macrophages, then stimulated with LPS, IL-1b, poly(I:C), TNF-a or PGN for 8 h. (f,g) Immunoblot analysis of extracts from TRIM31þ /þ or

TRIM31� /� mouse peritoneal macrophages stimulated with LPS for 4 h, and then treated for various times with cycloheximide (CHX). NLRP3, ASC,

Caspase-1 and NLRC4 expression levels were quantitated by measuring band intensities using ‘ImageJ’ software. The values were normalized to actin (g).

**Po0.01. (mean and s.d. of three samples in g, Student’s t-test). (h) Immunoblot analysis of extracts from HEK293T cells transfected with Myc-NLRP3

and Flag-TRIM31 expression plasmid then treated with MG132 (10mM) or chloroquine (10 mM) for 4 h. (i) Immunoblot analysis of extracts from HEK293T

cells transfected with Myc-NLRP3 and Flag-TRIM31 expression plasmid then treated with 3-MA as indicated for 4 h. (j) Immunoblot analysis of NLRP3

expression from TRIM31þ /þ or TRIM31� /� mouse peritoneal macrophages stimulated with LPS for 4 h, together with DMSO or MG132 (10mM)

treatment for 4 h. Similar results were obtained in three independent experiments.




plasmid (Fig. 4b), indicating TRIM31 mediated K48-linkedpolyubiquitination of NLRP3.

Under physiological conditions, the endogenous NLRP3 wasobserved to be robustly ubiquitinated with both K48 and K63linkage upon LPS stimulation (Fig. 4c). However, only K48-linkedpolyubiquitination of NLRP3 was almost completely abolished inTRIM31 deficient macrophages, and K63-linked ubiquitinationof NLRP3 was not affected (Fig. 4c). To determine whetherTRIM31 directly modified NLRP3, we reconstituted the NLRP3-ubiquitination reaction in vitro. We prepared recombinantNLRP3, recombinant TRIM31 (wt and DRING), purified E2enzyme UbcH5A, a mixture of E1 plus ubiquitin (wt, K48 onlyand K63 only). We observed ubiquitination of NLRP3 only whenTRIM31 wt was present (Fig. 4d). TRIM31 DRING lost the abilityof promoting NLRP3 ubiquitination. Furthermore, we foundubiquitination of NLRP3 was mediated by K48 linkage but not byK63 linkage (Fig. 4d). Taken together, these data indicated thatTRIM31 directly induced the K48-linked ubiquitination ofNLRP3 through its E3 ligase activity.

K48-linked polyubiquitination leads to the degradation oftarget proteins through 26S proteasome28. TRIM31 promotedproteasomal degradation of NLRP3 (Fig. 2g). However,TRIM31 mutants DRING and C16AC36A with the deprivationof the E3 ubiquitin ligase activity lost the ability to promotedegradation of NLRP3 (Fig. 4e–g). All together, these datademonstrated that TRIM31 mediated K48-linked ubiqutinationand proteasomal degradation through its E3 ubiquitin ligaseactivity.

LPS and IL-1b induce TRIM31 expression. Having establishedthe inhibitory role of TRIM31 in NLRP3 inflammasomeactivation, we next examined the expression status of TRIM31during the process of NLRP3 inflammasome activation. LPS or

IL-1b stimulation greatly induced TRIM31 expression at bothmRNA and protein level (Fig. 5a,b).

TRIM31 deficiency enhances IL-1b secretion in vivo. We nextinvestigated the biological effects of TRIM31 on NLRP3inflammasome activation in vivo. The induction of IL-1b byintraperitoneal (i.p.) injection of LPS is NLRP3 dependent29,30.We then examined whether TRIM31 deficiency could regulatethis induction of IL-1b. TRIM31 deficiency enhanced serumconcentrations of IL-1b but did not considerably increase theamount of TNF-a and IL-6 (Fig. 6a), indicating that TRIM31could specifically attenuate NLRP3 inflammasome activationin vivo.

TRIM31 deficiency aggravates alum-induced peritonitis. Wefurther investigated the effects of TRIM31 in an IL-1-dependentmouse peritonitis model31. Mouse peritonitis was induced by i.p.injection of alum. Peritoneal exudate cells (PECs) were collectedand alum-induced recruitment of inflammatory cells wasanalysed by flow cytometry. Alum challenge greatly upregulatedcaspase-1 cleavage in the lavage fluid and that this wassignificantly enhanced by TRIM31 deficiency (Fig. 6b), whichwas consistent with the observation in vitro. In addition, moreNLRP3 expression was also observed in the PECs lysates fromTRIM31 deficient mice (Fig. 6b). We consequently analysedalum-induced recruitment of inflammatory cells in the lavagefluid. Alum-induced recruitment of neutrophils and Ly6Cþ

monocytes was strongly enhanced in the TRIM31 deficientmice (Fig. 6c). Collectively, these findings further demonstratedthat TRIM31 could inhibit NLRP3 inflammasome activity andsubsequent immune cell accumulation in mouse peritonitisin vivo.

d

g h

e

f

Flag-TRIM31

Flag-TRIM31 C126

Flag-TRIM31 C16AC36A

Flag-TRIM31 C270Flag-TRIM31 N131Flag-TRIM31 N162

Flag-TRIM31 ΔRINGFlag-TRIM31 WT

Myc-NLRP3 WTMyc-NLRP3 ΔLRR

Myc-NLRP3 ΔNACHTMyc-NLRP3 ΔPYD

Myc-NLRP3

WT

ΔLRR

LRR

LRR

LRR

ΔNACHT

NACHT

NACHT

NACHTΔPYD

PYD

PYD

PYD

+

+ ++–

––

––

––

– –––

–

––––

+ + +

++

+

+ + + + + +

++

++

++

++–– –

– – – – ––––

–––––

–––––

–––––

– – ––– –

––– –

––

–– –

––––

––––

WT

C126

C270

N131

N162

ΔRING

RING B-box CC

CC

CC +

+

+

+

–

–

NLRP3 binding

CC

CC CC

CC

CC

B-box

B-box

B-box

RING

RING

Myc

Myc

Myc

Myc

Input

Input

Flag

Flag

TRIM31-GFP

TRIM31binding

Merge

20 μm

NLRP3-Myc

100 kDa

100 kDa

100 kDa

100 kDa

40 kDa

40 kDa

35 kDa

25 kDa

15 kDa

+

+

+

–

cFlag-TRIM31Myc-NLRP3

Myc-Caspase-1Myc-ASC

IP:Myc

Myc

*Input

Flag

Flag

100 kDa

40 kDa

40 kDa

25 kDa

40 kDa

+ ++

+

++

+––

––– –

–– –

IP:Flag

IP:Flag

Inpu

t

b

TRIM31

TRIM31

NLRP3

LPS (h) 0 4 8 8

100 kDa

40 kDa

40 kDa

+ + ++– – ––

IP:IgGIP:NLRP3

Flag-TRIM31

aNLRP3

+ ++

100 kDa

100 kDa

40 kDa

–

IP:Flag

IB:FlagInput

IB:NLRP3

IB:NLRP3

Figure 3 | TRIM31 interacts with NLRP3. (a) NLRP3 and Flag-tagged TRIM31 were obtained by in vitro transcription and translation. Interaction between

TRIM31 and NLRP3 was assayed by mixing TRIM31 and NLRP3 together followed by IP with Flag antibody and immunoblot analysis with NLRP3 antibody.

(b) Co-immunoprecipitation of endogenous TRIM31 with endogenous NLRP3 from mouse peritoneal macrophages stimulated with LPS for indicated time

periods. (c) HEK293Tcells expressing Flag-TRIM31 and Myc-NLRP3, Myc-Caspase-1 or Myc-ASC were lysed. Co-immunoprecipitation of Flag-TRIM31 with

Myc-NLRP3 from HEK293T cells. *Non-specific band. (d) HEK293T cells transfected with GFP-TRIM31 and Myc-NLRP3 were fixed and incubated with a

secondary antibody conjugated to Alexa Fluor 568. Colocalization between TRIM31 and NLRP3 was examined by Confocal microscopy. (e) Schematic

diagram of TRIM31 and its truncation mutants. (f) Flag-tagged TRIM31 or its mutants and Myc-NLRP3 were individually transfected into HEK293T cells.

The cell lysates were immunoprecipitated with an anti-Flag antibody and then immunoblotted with the indicated antibodies. (g) Schematic diagram of

TRIM31 and its truncation mutants. (h) Myc-tagged NLRP3 or its mutants and Flag-TRIM31 were individually transfected into HEK293T cells. The cell

lysates were immunoprecipitated with an anti-Flag antibody and then immunoblotted with the indicated antibodies. Similar results were obtained in three

independent experiments.




TRIM31 deficiency ameliorates DSS-induced colitis. DSS modelis one of the most extensively used to investigate innate immunemechanisms of colitis32. NLRP3 inflammasome is criticallyinvolved in the maintenance of intestinal homeostasis andprotection against colitis33. Besides spleen, TRIM31 expressed

extremely high in gut, including small intestine and colon(Supplementary Fig. 5), indicating the functions of TRIM31 inmaintaining intestinal homeostasis. We thus examined thepotential roles of TRIM31 in the DDS-induced colitis model.To study the contribution of TRIM31 to the development of

HA-Uba

d

e f g

b c

HA-Ub VectorTRIM31+/+ TRIM31–/–

+ ++– + –– +

– –+ + + + + +K48 K63

LPS

Ub

Ub(K48)

IP:NLRP3

InputNLRP3

Actin

Ub(K48)

–

–– – –

–

++ +++

– ++ +

+

Myc-NLRP3

Myc-NLRP3Flag-TRIM31 WT

Flag-TRIM31Flag-TRIM31 C16AC36A

HA

HA

170 kDa130 kDa

100 kDa

170 kDa130 kDa

100 kDa

170 kDa130 kDa

100 kDa

100 kDa

40 kDa

170 kDa

130 kDa

170 kDa

130 kDa

170 kDa

130 kDa

100 kDa

100 kDa

100 kDa

40 kDa

40 kDa

100 kDa

100 kDa

40 kDa

40 kDa

100 kDa

170 kDa130 kDa

100 kDa

100 kDa

40 kDa40 kDa

100 kDa

40 kDa

40 kDa

40 kDa

35 kDa

Ub(K48)

Myc

Myc-NLRP3Flag-TRIM31 WT

Flag-TRIM31 ΔRING TRIM31 ΔRING

Myc

Flag

Actin

Flag-TRIM31 C16AC36A TRIM31 WT TRIM31 WT

LPS

TRIM31 C16AC36ANLRP3

Actin

NLRP3

Actin

+ + + +– + – –– – + – – – – –+ +

– – – – +–+–– – – +

FlagActin

Myc

Flag

ActinInput

Ub wt K48 K63E1 + + + + + + + + + ++ +

+ ++ +

+ + +

+ ++ +

++ ++ +

+ +wt ΔRING wt ΔRING wt ΔRING

+

+–

+ + +– – – – –

+E2-UbcH5AE3-TRIM31NLRP3

NLRP3

TRIM31 wt

TRIM31 C58

Input

IP:MycIP:Myc

Figure 4 | TRIM31 promotes K48-linked polyubiquitination of NLRP3. (a) Immunoblot analysis of lysates from HEK293T cells transfected with

HA-tagged ubiquitin (HA-Ub), Myc-NLRP3 and TRIM31 WT or C16AC36A, followed by IP with anti-Myc, probed with anti-HA or K48-Ub. (b) Immunoblot

analysis of lysates from HEK293T cells transfected with HA-tagged K48-linked ubiquitin (K48-Ub) or HA-tagged K63-linked ubiquitin (K63-Ub),

Myc-NLRP3 and TRIM31, followed by IP with anti-Myc, probed with anti-HA. (c) Immunoblot analysis of lysates from TRIM31þ /þ or TRIM31� /� mouse

peritoneal macrophages, followed by IP with anti-NLRP3, probed with anti-Ub, K48-Ub or K63-Ub. (d) In vitro ubiquitination assay was performed in the

presence of Ub (wt, K48 or K63), E1, E2-UbcH5A, NLRP3 and TRIM31 (wt or DRING mutant). The ubiquitination of NLRP3 was examined with NLRP3

antibody. (e) Immunoblot analysis of extracts from HEK293T cells transfected with Myc-NLRP3 and Flag-tagged TRIM31 or its mutants. (f) Immunoblot

analysis of extracts from THP-1 cells transfected with TRIM31 wild type (WT) or TRIM31 C16AC36A. (g) Immunoblot analysis of extracts from THP-1 cells

transfected with TRIM31 WT or TRIM31 DRING then stimulated with LPS for 4 h. Similar results were obtained in three independent experiments.

10

a b

8

8 12 24

6

4

42

2

TR

IM31

mR

NA

(fo

ld)

TR

IM31

mR

NA

(fo

ld)

Actin

Actin

0

8

6

4

2

00 8 12 24420IL-1β (h)

IL-1β (h) 8 12 24420

8 12 24

40 kDa

40 kDa

40 kDa

40 kDa

40

LPS (h)

LPS (h)

TRIM31

TRIM31

Figure 5 | IL-1b and LPS induce TRIM31 expression. (a,b) RT-PCR analysis (a) or immunoblot analysis (b) of TRIM31 expression from mouse peritoneal

macrophages stimulated with LPS or IL-1b for various times. Similar results were obtained in three independent experiments. Data are representative of

three experiments (mean and s.d. of six samples in a).




colitis, we first assessed the clinical features of age- andsex-matched WT and TRIM31 deficient mice after oraladministration of 3% DSS in drinking water. TRIM31� /�

mice suffered from less body weight loss from day 4 on(Fig. 7a). Simultaneously, rectal bleeding scores and stoolconsistency scores of TRIM31� /� mice were significantlyattenuated compared with those of DSS-fed WT mice

(Fig. 7b,c). To further assess the severity of colitis, colon lengthwas measured in DSS-fed WT and TRIM31� /� mice. Colons ofTRIM31� /� mice were longer than those of WT miceadministered with DSS (Fig. 7d,e). The severity of colonicinflammation and ulceration was further evaluated byhistopathological analysis using haematoxylin & eosin (H&E)staining. DSS-induced mucosal damage was characterized by

1,200

1,000

**

800

600

0 0

0 0

0.5

1

1.5

2**

1

Neu

trop

hils

(×

106 )

Ly6C

+ m

onoc

ytes

(×

106 )

2

3cb

a

**

10

20

30

40

50

60

70 TRIM31+/+

TRIM31–/–

TRIM31+/+

TRIM31+/+

TRIM31–/–TRIM31–/–

2

4

6

8

400

IL-1

β (p

g m

l–1)

IL-6

(ng

ml–1

)

TN

F-α

(ng

ml–1

)

200

Alum

Alum Alum AlumAlum

130 kDa100 kDa

40 kDa

40 kDa

25 kDa

15 kDa

25 kDa– – + + + + + + + +

p20

NLRP3

Caspase-1

ASC

Actin

0

PBSPBS

LPS

LPS

PBS

PBSPBS

PBSLP

SLP

SPBS

PBSLP

SLP

S

Figure 6 | TRIM31 deficiency enhances IL-1b secretion and aggravates Alum-induced peritonitis in vivo. (a) ELISA analysis of serum levels of IL-1b,

TNF-a and IL-6 from TRIM31þ /þ or TRIM31� /� mice after i.p. LPS injection. (b,c) TRIM31þ /þ or TRIM31� /� deficiency mice were i.p. injected with

alum for 12 h. PECs were lysed and analysed for the expression of Caspase-1, NLRP3 and ASC by immunoblot (b). Absolute numbers of neutrophils or

Ly6Cþ monocytes recruited to the peritoneum were analysed by fluorescence-activated cell sorting (five mice per group) (c). **Po0.01. (Student’s t-test).

Data are representative of three experiments (mean and s.d. of four to five samples in a and c).

110a

f g

h

b c d e

**

****

******

TRIM31+/+

TRIM31+/+

TRIM31+/+ TRIM31

+/+

TRIM31+/+

TRIM31+/+

TRIM31+/+

TRIM31–/–

TRIM31–/–

TRIM31–/–

TRIM31–/–

TRIM31–/–

TRIM31–/–

TRIM31–/–

TRIM31

NLRP3

NLRP3

Caspase-1

Actin

ASC

Actin

100

90

80

Bod

y w

eigh

t (%

)

70

60 0 0

0.5 0.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0***

1 11.5

1.52

2

Ble

edin

g sc

ore

Sto

ol s

core

Col

on le

ngth

(cm

)

2.52.53

33.5

0

Time after DSS-administration(days)



1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6

DSS

55 kDa

55 kDa

25 kDa

100 kDa

100 kDa

40 kDa

40 kDa

Figure 7 | TRIM31 deficiency ameliorates DSS-induced colitis. Mice were given 3% DSS in their drinking water for 5 days, followed by regular

drinking water. (a) Body weight, (b) stool consistency and (c) rectal bleeding score were scored daily. (d–h) Mice were killed on day 6. Macroscopic

appearances (d) and colon lengths (e) of the mice were measured. Histopathological changes in colon tissue were examined by H&E staining (f) Scale

bars, 50mm. Immunoblot analysis of lysates from colon tissue (g,h). *Po0.05, **Po0.01, ***Po0.001 (one-way analysis of variance, ANOVA).




crypt loss and infiltrating leucocytes. Colonic sections of DSS-fedWT mice displayed severe transmural inflammation with focalareas of extensive ulceration and necrotic lesions (Fig. 7f).Inflammatory infiltrates filled the lamina propria and submucosain areas where the mucosa was intact and often effaced thenormal architecture of the tissue (Fig. 7f). Strikingly, colonicsections of DSS-fed TRIM31� /� mice exhibited lessinflammatory cell infiltration and intact colonic architecturewith no apparent ulceration (Fig. 7f).

To confirm the roles of TRIM31 in the model, we examinedTRIM31 and NLRP3 expressions in colon tissues. In WT mice,TRIM31 constitutely expressed in colon tissues (Fig. 7g,Supplementary Fig. 5), and that maintains NLRP3 proteinexpression in a relatively low level (Fig. 7g). Following DSStreatment, TRIM31 was downregulated and NLRP3 proteinexpression was elevated (Fig. 7g). In DSS-administeredTRIM31� /� mice, NLRP3 protein expression in colon tissueswas further enhanced (Fig. 7h). Collectively, these data indicatedthat TRIM31 played crucial roles in limiting NLRP3 expressionand thus exacerbated DSS-induced colitis.

DiscussionUbiqutination is a crucial post-transcriptional modification tomodulate the intensity of innate immune responses, includingregulation of NLRP3 inflammasome activity. For example,K63-linked ubiquitination of ASC targets the AIM2–ASCinflammasome complex for destruction in autophagosomes34,35.The linear ubiquitin assembly complex-mediated ASC linearubiquitination is required for NLRP3 inflammasome activation36.A20 restricts ubiquitination of pro-interleukin-1b proteincomplexes to suppress NLRP3 inflammasome activity37.TRIM33 ubiquitinates DHX33 and is essential for thecytosolic RNA-induced NLRP3 inflammasome activation38. Thedeubiquitination of NLRP3 by the BRCC3 complex is requiredfor NLRP3 inflammasome activation13,39,40. E3 ubiquitin ligaseMARCH7 promoted NLRP3 K48-linked ubiquitination andsubsequent degradation in an autophagy-dependent pathway18.However, whether the ubiquitinated NLRP3 could be degradedvia a ubiquitin–proteasome pathway is still unknown. In thepresent study, we found that NLRP3 could be ubiquitinated withK48 linkage in both resting and activated macrophages, and E3

ubiquitin ligase TRIM31 mediated this modification. Thus,TRIM31 promoted proteasomal degradation of NLRP3 andthen inhibited the NLRP3 inflammasome activation. Collectively,several E3 ubiquitin ligases, such as TRIM31, MARCH7, DHX33and so on, could negatively or positively regulate NLRP3inflammasome activation at different levels and via differentmechanisms. As for the regulation of NLRP3 expression, E3ubiquitin ligase TRIM31 and MARCH7 could promote thedegradation of NLRP3 via a ubiquitin–proteasome pathwayand autophagy-dependent pathway, respectively. Thus, the E3ubiquitin ligases could work together to fine tune the activation ofNLRP3 inflammasome to maintain immune homeostasis. Ourresearch identify TRIM31 as a physiological regulator of NLRP3expression, and that will further explain the complexities ofhomeostatic control of innate immune signalling.

Based on the experimental data, we propose a model toillustrate how TRIM31 feedback negatively regulates NLRP3inflammasome activation (Fig. 8). In resting macrophages,the constitutively expressed TRIM31-mediated proteasomaldegradation of NLRP3 protein to maintain its low expressionand limit the NLRP3 inflammasome assembly (Fig. 8a). FollowingLPS stimulation, NLRP3 expression is induced rapidly throughNF-kB activation, which could facilitate NLRP3 inflammasomeassembly and activation. Simultaneously, TRIM31 expression isinduced by LPS stimulation and IL-1b secretion. TRIM31 thendirectly bound to NLRP3 and promoted its K48-linkedubiquitination, leading to its proteasomal degradation (Fig. 8b).Thus, the activation of NLRP3 inflammasome was limited in aproper intensity and time course to avoid detrimental effects.

Improper NLRP3 inflammasome activation has beenlinked to a variety of diseases5, such as infectiousdiseases, autoinflammatory and autoimmune diseases. NLRP3inflammasome activity is critical for host response to microbialpathogens. Excessive NLRP3 inflammasome activation playscrucial causative or contributing roles in the initial ofautoinflammatory and autoimmune diseases5. However, optimalNLRP3 inflammasome activation is beneficial for someinflammatory disorders, such as inflammatory bowel diseases33.Thus, a better understanding of the balance between beneficialand detrimental NLRP3 inflammasome activation is needed.

In the present study, we found that TRIM31 constitutelyexpressed in macrophages and negatively regulated NLRP3

TLR4

NF-κB NLRP3

NLRP3

NLRP3degradation Pro-Caspase-1

ASCTRIM31

UbUb

UbUb

Ub

TLR4

LPS IL-1β

Pro-IL-1β

IL-1R

NF-κB

NF-κB

NLRP3

NLRP3degradation Caspase-1

Pro-Caspase-1

TRIM31 TRIM31

UbUb

UbUb

Ub

ASC

a b

Unactivated phase Activated phase

Figure 8 | Working model for TRIM31 inhibiting NLRP3 inflammasome activation. (a) In resting macrophages, the constitutively expressed TRIM31 binds

to NLRP3, promotes K48-linked ubiquitination and proteasomal degradation of NLRP3, and maintains its low expression. (b) Following NLRP3

inflammasome activation, TRIM31 expression is markedly induced by LPS and IL-1b, resulting in the inhibition of NLRP3 expression and subsequent

inflammasome activation.




inflammasome activation to prevent unwanted activation in bothresting and activation state. In vivo experiments also confirmedthe beneficial roles of TRIM31 in the alum-induced peritonitis,an NLRP3 dependent acute inflammatory model. Furthermore,TRIM31 also constitutively expressed in gut, limited NLRP3expression in a relatively low level and thus possessed potentialfunctions in maintaining intestinal homeostasis. However,TRIM31þ /þ mice were more susceptible to acute DSS-inducedcolitis, compared with TRIM31� /� mice. Recent evidence hasindicated that NLRP3 confers protection against acutecolitis33,41,42. TRIM31 expression was downregulated in theDSS-induced colitis model, and NLRP3 expression wasupregulated accordingly. Thus, NLRP3 could exert itsprotection roles when colitis was initiated. In TRIM31� /�

mice, NLRP3 expression was further increased in the colon ofDSS-induced colitis model. So, TRIM31 deficiency amelioratesDSS-induced colitis.

In summary, we identified TRIM31 as a feedback suppressorof NLRP3 inflammasome activity. Because of the conservedbiological function of NLRP3 inflammasome in multiplephysiological processes, fine-tuning of its activity is critical forthe disease resistance and maintenance of immune homeostasis.Our research provides an explanation on the limitation of NLRP3inflammasome activation under physiological conditions.Furthermore, TRIM31 may provide a useful therapeutic targetfor the intervention of diseases with improper inflammasomeactivation, such as autoinflammatory diseases, colitis, obesity anddiabetes.

MethodsMice. TRIM31� /� mice on a C57BL/6J background were generated by CyagenBiosciences Inc. (Guangzhou, China) using TALEN technology. A pair of TALENconstructs for TRIM31 knockout (KO) were cloned into a mammalian expressionvector pCMV-TALEN and capped, polyA-tailed mRNA for injection wereproduced using the Ambion mMessage mMachine kit. The KO mice wereproduced by microinjecting TALEN mRNAs into fertilized eggs from C57BL/6strain. The KO alleles have been sequence validated. Three kinds of KO mice weregenerated and named as KO1, KO2 and KO3. KO1 has one missing base pairs(C, 97 of the ORF), KO2 has one missing base pairs (G, 94 of the ORF) and KO3has five missing base pairs (GGGCA, 94-98 of the ORF). All the three KOs caused aframe shift. The mRNA transcribed from targeted allele with frame shift undergoesnonsense-mediated decay. KO1 mice were used in most experiments if nospecial noted. The mice were backcrossed with C57BL/6 mice for more than sixgenerations. C57BL/6 mice were obtained from Vital River Laboratory AnimalTechnology Co. (Beijing, China). All animal experiments were undertaken inaccordance with the National Institute of Health Guide for the Care and Use ofLaboratory Animals, with the approval of the Scientific Investigation Board ofMedical School of Shandong University (Jinan, Shandong Province, China).

Reagents. ATP, MG132, chloroquine, cycloheximide, phorbol myristateacetate, LPS (Escherichia coli, 055:B5), peptidoglycan (PGN), poly(I:C), 3-MA,anti-TRIM31, anti-Myc and anti-Flag were from Sigma-Aldrich (St Louis, MO);Dextran sulfate sodium (DSS, 36-50 kDa) was bought from MP Biomeicals(Aurora, OH). Imject Alum was from Thermo Scientific; IL-1b and TNF-a werefrom PeproTech (Rocky Hill, NJ); Flagellin and poly(dA:dT) were from Invivogen(San Diego, CA). Nigericin, anti-caspase-1 p45&p20, anti-NLRP3 and anti-ASCwere from AdipoGen; anti-hemagglutinin (HA), anti-Ub, anti-b-actin, protein Gagarose used for IP and horseradish peroxidase-conjugated secondary antibodieswere from Santa Cruz Biotechnology (Santa Cruz, CA); anti-IL-1b p31&p17,anti-p-JNK, anti-p-p38, anti-p-ERK, anti-JNK, anti-p38, anti-ERK1/2, anti-p-IkBa,anti-K48-ub and anti-K63-ub were from Cell Signaling Technology (Beverly, MA);anti-AIM2 and anti-NLRC4 were from Abcam (Cambridge, MA).

Cell culture. To obtain mouse primary peritoneal macrophages, C57BL/6J mice(female, 4–6 weeks old) were injected intraperitoneally with 3% Brewer’s thiogly-collate broth. Three days later, PEC were harvested and incubated. Two hours later,nonadherent cells were removed and the adherent monolayer cells were used asperitoneal macrophages14,40,43. Human THP-1 and human embryonic kidney(HEK293T) cells were obtained from American Type Culture Collection(Manassas, VA). Phorbol myristate acetate-activated THP-1 cells were used ashuman macrophages. The cells were cultured at 37 �C under 5% CO2 in DMEMsupplemented with 10% FCS (Invitrogen-Gibco), 100 U ml� 1 penicillin, and100mg ml� 1 streptomycin. The concentration of agonists or stimuli were used as

below: LPS 100 ng ml� 1 for mouse primary peritoneal macrophages, LPS1 mg ml� 1 for THP-1 cells, ATP 5 mM, nigericin 50 mM and alum 200 mg ml� 1,IL-1b 10 ng ml� 1, poly(I:C) 10 mg ml� 1, TNF-a 200 ng ml� 1 and PGN5 mg ml� 1. Poly(dA:dT) and flagellin were transfected into macrophages with thefinal concentration as below: poly(dA:dT) 200 ng ml� 1 and flagellin 200 ng ml� 1.

Plasmids construction and transfection. TRIM31 expression plasmid was con-structed by PCR-based amplification of cDNA from THP-1 cells, and then clonedinto the pFLAG-CMV-2 eukaryotic expression vector (Sigma-Aldrich). Caspase-1expression plasmid was constructed by PCR-based amplification of cDNA fromTHP-1 cells, and then cloned into the pCMV-Myc eukaryotic expression vector(Beyotime, China). Myc-NLRP3 and Myc-ASC plasmids were provided by Dr JohnC. Reed (Sanford-Burnham Medical Research Institute, La Jolla, CA). Expressionvectors for HA-Ub WT, mutant K48 and mutant K63 were from Dr Hui Xiao(Institut Pasteur of Shanghai, CAS, Shanghai, China). MyD88 plasmid was giftsfrom Dr Xuetao Cao (Second Military Medical University, Shanghai, China).NF-kB reporter plasmid was purchased from Stratagene. Truncated mutants ofTRIM31 or NLRP3 were generated using the KOD-Plus-Mutagenesis kit (Toyobo,Osaka, Japan). All constructs were confirmed by DNA sequencing. Plasmids weretransiently transfected into HEK293T cells or THP-1 cells with jetPRIME reagents(Polyplus) according to the manufacturer’s instructions.

RNA interference assay. siRNAs were synthesized as following: murine TRIM31:50-GCUCACUAAAUCCUUGAAA-30 , human TRIM31: 50-GGACCA-CAAAUCCCAUAAU-30 and negative control: 50-UUCUCCGAACGUGU-CACGU-30 . These siRNA duplexes were transfected into mouse peritonealmacrophages or THP-1 cells using INTERFERin reagents (PolyPlus) according tothe manufacturer’s instructions.

ELISA. The concentrations of mouse IL-1b, mouse TNF-a and mouse IL-6 weremeasured using ELISA kits (Dakewe Biotech Company Ltd., Shenzhen, China)according to the manufacturer’s instruction.

RNA quantitation. Total RNA was extracted with TRIzol reagent according to themanufacturer’s instructions (Invitrogen). A LightCycler (ABI PRISM 7000) and aSYBR RT-PCR kit (Takara) were used for quantitative real-time RT-PCR analysis.Specific primers used for RT-PCR assays were the sequences of primers used forRT-PCR were 50-TGGATGGGTTTGCTGGGAT-30 , 50-CTGCGTGTAGCGACTGTTGAG-30 for NLRP3; 50-ACCTTCCAGGATGAGGACATGA-30 , 50-AACGTCACACACCAGCAGGTTA-30 for IL-1b; 50-CCAGAGTCAAACCGTGAGCG-30 ,50-GGCAACTTGGAGCCCGAA-30 for TRIM31; 50-GCCACCACGCTCTTCTGTCT-30 , 50-TGAGGGTCTGGGCCATAGAAC-30 for TNF-a; and 50-TGTTACCAACTGGGACGACA03, 50-CTGGGTCATCTTTTCACGGT-30 for b-actin. Dataare normalized to b-actin expression in each sample.

Immunoprecipitation and immunoblot analysis. For IP, whole-cell extracts werelysed in IP buffer containing 1.0% (vol/vol) Nonidet P 40, 50 mM Tris–HCl,pH 7.4, 50 mM EDTA, 150 mM NaCl, and a protease inhibitor ‘cocktail’ (Merck).After centrifugation for 10 min at 14,000 g, supernatants were collected andincubated with protein G Plus-Agarose Immunoprecipitation reagent together withspecific antibody. After 6 h of incubation, beads were washed five times with IPbuffer. Immunoprecipitates were eluted by boiling with 1% (wt/vol) SDS samplebuffer. For immunoblot analysis, cells were lysed with M-PER Protein ExtractionReagent (Pierce, Rockford, IL) supplemented with a protease inhibitor ‘cocktail’,then protein concentrations in the extracts were measured with a bicinchoninicacid assay (Pierce, Rockford, IL). Equal amounts of extracts were separated bySDS–polyacrylamide gel electrophoresis (PAGE), and then were transferred ontonitrocellulose membranes for immunoblot analysis. Cell culture supernatantswere harvested and concentrated for immunoblot with Amicon Ultra 10 K fromMillipore. The concentration of primary antibodies used for immunoblot was1 mg ml� 1. The secondary antibodies were used 1 in 5,000. Uncropped scans ofimmunoblots are provided as Supplementary Figs 6–8.

In vitro binding and ubiquitination assay. NLRP3, TRIM31 WT and DRINGmutant proteins were expressed with a TNT Quick Coupled Transcription/Translation System (Promega) according to the instructions of the manufacturer22.Binding assays were performed by mixing Flag-tagged TRIM31 and NLRP3together, followed by IP with Flag antibody and WB with NLRP3 antibody.Ubiquitination was analysed with an ubiquitination kit (Boston Biochem)following protocols recommended by the manufacturer.

Luciferase assay. Luciferase activities were measured with Dual-LuciferaseReporter Assay System (Promega) on a microplate luminometer (Centro LB 960;Berthold, Wildbad, Germany) according to the manufacturer’s instructions. Dataare normalized for transfection efficiency by subtracting Firefly luciferase activitywith that of Renilla luciferase.




Immunofluorescence staining and confocal analysis. HEK293T cells transientlytransfected with plasmids encoding Myc-NLRP3 and GFP-TRIM31 were culturedfor 24 h. Myc-tagged NLRP3 was detected directly following fixation and washing.Anti-Myc Ab was used at 1:1,000 dilutions in the blocking solution. The cells werethen incubated with Alexa Fluor 568-conjugated secondary Ab (Molecular Probes,Invitrogen) diluted 1:1,000 in blocking solution. Nuclei were stained with DAPI(40 , 60-diamidino-2-phenylindole hydrochloride; Molecular Probes, Invitrogen).Cells were examined with confocal laser microscopy (LSM780, Carl Zeiss,Oberkochen, Germany).

In vivo LPS challenge. WT or TRIM31 deficiency mice (females, 6 weeks old) werei.p. injected with 10 mg kg� 1 LPS or PBS30. After 2 h mice were killed, and serumlevels of IL-1b, TNF-a and IL-6 were measured by ELISA.

In vivo peritonitis. WT or TRIM31 deficiency mice (females, 6 weeks old) were i.p.injected with 700mg alum for 12 h. Peritoneal cavities were washed with 6 ml ofPBS. The peritoneal fluids were harvested and concentrated for ELISA analysis withAmicon Ultra 10 K from Millipore. PECs were analysed by FACS14,31.

DSS-induced colitis. Colitis was induced in WT or TRIM31 deficiency C57BL/6mice (females, 6 weeks old) with 3% (wt/vol) DSS dissolved in drinking water for 5days, followed by regular drinking water. Mice were monitored for body weight,stool consistency and the presence of occult blood every day. Scoring for stoolconsistency and occult blood was done as described previously32,33. In brief, stoolscores were determined as follows: 0, well-formed pellets; 1, semiformed stools thatdid not adhere to the anus; 2, semiformed stools that adhered to the anus; 3, liquidstools that adhered to the anus. Bleeding scores were determined as follows: 0, noblood as tested with hemoccult (Beckman Coulter); 1, positive hemoccult; 2, bloodtraces in stool visible; 3, gross rectal bleeding. After day 6, the entire colon wasexcised to measure the length of the colon and the weight of caecum. ThenB0.5 cm of mice colon tissues close to the rectum was collected for histologicalanalysis and protein extraction. For histological analysis, formalin-fixed andparaffin-embedded segments of colon tissue were sectioned at 4 mm in thickness,and the sections were stained with hematoxylin and eosin (H&E). TRIM31,NLRP3, Caspase-1 and ASC expression were examined with immunoblot analysis.

Statistical analysis. All experiments were independently performed three times.Data are presented as means±s.d. of three or four experiments. Analysis wasperformed using a Student’s t-test or One-way analysis of variance (ANOVA).Values of Po0.05 were considered to be statistically significant.

Data availability. The data that support this study are available within the articleand its Supplementary Information files or available from the authors uponrequest.

References1. Schroder, K. & Tschopp, J. The inflammasomes. Cell 140, 821–832 (2010).2. Davis, B. K., Wen, H. & Ting, J. P. The inflammasome NLRs in immunity,

inflammation, and associated diseases. Annu. Rev. Immunol. 29, 707–735(2011).

3. Rathinam, V. A., Vanaja, S. K. & Fitzgerald, K. A. Regulation of inflammasomesignaling. Nat. Immunol. 13, 333–342 (2012).

4. Latz, E., Xiao, T. S. & Stutz, A. Activation and regulation of the inflammasomes.Nat. Rev. Immunol. 13, 397–411 (2013).

5. Guo, H., Callaway, J. B. & Ting, J. P. Inflammasomes: mechanism of action, rolein disease, and therapeutics. Nat. Med. 21, 677–687 (2015).

6. Strowig, T., Henao-Mejia, J., Elinav, E. & Flavell, R. Inflammasomes in healthand disease. Nature 481, 278–286 (2012).

7. Wen, H., Ting, J. P. & O’Neill, L. A. A role for the NLRP3 inflammasome inmetabolic diseases--did Warburg miss inflammation? Nat. Immunol. 13,352–357 (2012).

8. Choi, A. J. & Ryter, S. W. Inflammasomes: molecular regulation andimplications for metabolic and cognitive diseases. Mol. Cells 37, 441–448(2014).

9. Broderick, L. et al. The inflammasomes and autoinflammatory syndromes.Annu. Rev. Pathol. 10, 395–424 (2015).

10. Dowling, J. K. & O’Neill, L. A. Biochemical regulation of the inflammasome.Crit. Rev. Biochem. Mol. Biol. 47, 424–443 (2012).

11. Haneklaus, M., O’Neill, L. A. & Coll, R. C. Modulatory mechanisms controllingthe NLRP3 inflammasome in inflammation: recent developments. Curr. Opin.Immunol. 25, 40–45 (2013).

12. Bauernfeind, F. G. et al. Cutting edge: NF-kappaB activating patternrecognition and cytokine receptors license NLRP3 inflammasome activation byregulating NLRP3 expression. J. Immunol. 183, 787–791 (2009).

13. Juliana, C. et al. Non-transcriptional priming and deubiquitination regulateNLRP3 inflammasome activation. J. Biol. Chem. 287, 36617–36622 (2012).

14. Huai, W. et al. Aryl hydrocarbon receptor negatively regulates NLRP3inflammasome activity by inhibiting NLRP3 transcription. Nat. Commun. 5,4738 (2014).

15. Bauernfeind, F. et al. NLRP3 inflammasome activity is negatively controlled bymiR-223. J. Immunol. 189, 4175–4181 (2012).

16. Haneklaus, M. et al. Cutting edge: miR-223 and EBV miR-BART15 regulate theNLRP3 inflammasome and IL-1b production. J. Immunol. 189, 3795–3799(2012).

17. Chuang, S. Y. et al. TLR-induced PAI-2 expression suppresses IL-1b processingvia increasing autophagy and NLRP3 degradation. Proc. Natl Acad. Sci. USA110, 16079–16084 (2013).

18. Yan, Y. et al. Dopamine controls systemic inflammation through inhibition ofNLRP3 inflammasome. Cell 160, 62–73 (2015).

19. Versteeg, G. A., Benke, S., Garcıa-Sastre, A. & Rajsbaum, R. InTRIMsicimmunity: positive and negative regulation of immune signaling by tripartitemotif proteins. Cytokine Growth Factor Rev. 25, 563–576 (2014).

20. Meyer, M., Gaudieri, S., Rhodes, D. A. & Trowsdale, J. Cluster of TRIM genesin the human MHC class I region sharing the B30.2 domain. Tissue Antigens61, 63–71 (2003).

21. Uchil, P. D. et al. TRIM protein-mediated regulation of inflammatory andinnate immune signaling and its association with antiretroviral activity. J. Virol.87, 257–272 (2013).

22. Wang, P., Zhao, W., Zhao, K., Zhang, L. & Gao, C. TRIM26 negatively regulatesinterferon-b production and antiviral response through polyubiquitination anddegradation of nuclear IRF3. PLoS Pathog. 11, e1004726 (2015).

23. Suzuki, M. et al. TRIM39 negatively regulates the NFkB-mediated signalingpathway through stabilization of cactin. Cell Mol. Life Sci. 73, 1085–1101 (2016).

24. Zhao, W., Wang, L., Zhang, M., Yuan, C. & Gao, C. E3 ubiquitin ligasetripartite motif 38 negatively regulates TLR-mediated immune responses byproteasomal degradation of TNF receptor-associated factor 6 in macrophages.J. Immunol. 188, 2567–2574 (2012).

25. Hu, M. M. et al. TRIM38 inhibits TNFa- and IL-1b-triggered NF-kB activationby mediating lysosome-dependent degradation of TAB2/3. Proc. Natl Acad. Sci.USA 111, 1509–1514 (2014).

26. Cai, X. et al. Tripartite motif containing protein 27 negatively regulates CD4T cells by ubiquitinating and inhibiting the class II PI3K-C2b. Proc. Natl Acad.Sci. USA 108, 20072–20077 (2011).

27. Zheng, Q. et al. Siglec1 suppresses antiviral innate immune response byinducing TBK1 degradation via the ubiquitin ligase TRIM27. Cell Res. 25,1121–1136 (2015).

28. Jiang, X. & Chen, Z. J. The role of ubiquitylation in immune defence andpathogen evasion. Nat. Rev. Immunol. 12, 35–48 (2011).

29. He, Y., Franchi, L. & Nunez, G. TLR agonists stimulate Nlrp3-dependent IL-1bproduction independently of the purinergic P2X7 receptor in dendritic cellsand in vivo. J. Immunol. 190, 334–339 (2013).

30. Coll, R. C. et al. A small-molecule inhibitor of the NLRP3 inflammasome forthe treatment of inflammatory diseases. Nat. Med. 21, 248–255 (2015).

31. Guarda, G. et al. Type I interferon inhibits interleukin-1 production andinflammasome activation. Immunity 34, 213–223 (2011).

32. Wirtz, S., Neufert, C., Weigmann, B. & Neurath, M. F. Chemically inducedmouse models of intestinal inflammation. Nat. Protoc. 2, 541–546 (2007).

33. Zaki, M. H. et al. The NLRP3 inflammasome protects against loss of epithelialintegrity and mortality during experimental colitis. Immunity 32, 379–391(2010).

34. Shi, C. S. et al. Activation of autophagy by inflammatory signals limits IL-1bproduction by targeting ubiquitinated inflammasomes for destruction. Nat.Immunol. 13, 255–263 (2012).

35. Shimada, K. et al. Oxidized mitochondrial DNA activates the NLRP3inflammasome during apoptosis. Immunity 36, 401–414 (2012).

36. Rodgers, M. A. et al. The linear ubiquitin assembly complex (LUBAC) isessential for NLRP3 inflammasome activation. J. Exp. Med. 211, 1333–1347(2014).

37. Duong, B. H. et al. A20 restricts ubiquitination of pro-interleukin-1b proteincomplexes and suppresses NLRP3 inflammasome activity. Immunity 42, 55–67(2015).

38. Weng, L. et al. The E3 ubiquitin ligase tripartite motif 33 is essential forcytosolic RNA-induced NLRP3 inflammasome activation. J. Immunol. 193,3676–3682 (2014).

39. Lopez-Castejon, G. et al. Deubiquitinases regulate the activity of caspase-1 andinterleukin-1b secretion via assembly of the inflammasome. J. Biol. Chem. 288,2721–2733 (2013).

40. Py, B. F., Kim, M. S., Vakifahmetoglu-Norberg, H. & Yuan, J. Deubiquitinationof NLRP3 by BRCC3 critically regulates inflammasome activity. Mol. Cell 49,331–338 (2013).

41. Allen, I. C. et al. The NLRP3 inflammasome functions as a negative regulator oftumorigenesis during colitis-associated cancer. J. Exp. Med. 207, 1045–1056(2010).

42. Hirota, S. A. et al. NLRP3 inflammasome plays a key role in the regulation ofintestinal homeostasis. Inflamm. Bowel Dis. 17, 1359–1372 (2011).




43. Zhang, M. et al. TRAF-interacting protein (TRIP) negatively regulates IFN-bproduction and antiviral response by promoting proteasomal degradation ofTANK-binding kinase 1. J. Exp. Med. 209, 1703–1711 (2012).

AcknowledgementsThis work was supported by Grants from the National Natural Science Foundation ofChina (81622030, 31370017, 31570867, 81525012 and 81471538), Shandong ProvincialNature Science Foundation for Distinguished Young Scholars (JQ201420) and Keyresearch and development program of Shandong Province (2015GSF118159).

Author contributionsW.Z. conceived the study, designed the experiments and provided overall direction; H.S.,B.L., W.H., Z.Y and W.W. did the experiments; J.Z. did the artwork of the model figure;C.G. provided TRIM31 deficient mice; L.H., G.J. and L.Z. provided expertise and advice;H.S., B.L., W.H. and W.Z analysed the data and wrote the paper.

Additional informationSupplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Competing financial interests: The authors declare no competing financial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

How to cite this article: Song, H. et al. The E3 ubiquitin ligase TRIM31 attenuatesNLRP3 inflammasome activation by promoting proteasomal degradation of NLRP3.Nat. Commun. 7, 13727 doi: 10.1038/ncomms13727 (2016).

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

This work is licensed under a Creative Commons Attribution 4.0International License. The images or other third party material in this

article are included in the article’s Creative Commons license, unless indicated otherwisein the credit line; if the material is not included under the Creative Commons license,users will need to obtain permission from the license holder to reproduce the material.To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

r The Author(s) 2016





http://npg.nature.com/reprintsandpermissions/

http://npg.nature.com/reprintsandpermissions/

http://creativecommons.org/licenses/by/4.0/


the e3 ubiquitin ligase trim31 attenuates nlrp3 inflammasome … · 2019-09-17 · ubiquitin ligase...

Documents